Easy RaspberryPi4 GUI system managment
-
Updated
Jan 27, 2026 - Python
Easy RaspberryPi4 GUI system managment
A set of utilities for monitoring and customizing GPU performance
Local AI server on Raspberry Pi 5: one-command installer, stable 3 GHz overclock, Ollama on LAN, Python mini-agent, optional Hailo AI HAT module. No cloud, no API keys.
A lightweight sensors monitor for your terminal
Mega Drive Overclock system
Underclock, environment friendly, your GPU will start to counting sheep.
⚙️ Android kernel compilation, custom kernels, KernelSU, governors, schedulers — complete reference for power users.
NVIDIA GPU overclock controller with watchdog, health monitoring, and driver recovery
To associate your repository with the overclock topic, visit your repo's landing page and select "manage topics."